fix(booth): put the blur toggle on every item kind, and make it look like a control

The operator asked "no UI option to blur/unblur?" and was right twice over.

MISSING ENTIRELY ON TWO OF THREE BRANCHES. booth.html renders docs, media and
everything-else through separate <figure> blocks. The toggle went into the
media branch only, so inline docs -- the branch that puts readable text straight
on the page, the one that needs blur most -- had no control at all, and `other`
files only got a caption row if they happened to carry a caption.

This is the SECOND time this feature shipped having patched some branches and
not others; the blurred class itself had the same gap one commit ago. So the
toggle is now a single Jinja macro called from all three sites, which makes
"patched two of three" impossible rather than merely unlikely, and
test_every_item_kind_gets_exactly_one_blur_toggle counts toggles against figures
across mixed kinds so a fourth branch cannot quietly skip it.

INVISIBLE WHERE IT DID RENDER. v1 was a bare `◌` at 0.78rem in --muted with no
border, no label and no hover affordance. It now reads `◌ blur` / `◉ blurred`
with a border, matching the other per-item controls. A control nobody can find
is a control that is not there.

Docs get it in the doc bar beside ⤢ ⬇ ✕, with stopPropagation so submitting it
does not collapse the <details> it lives inside.

Verified live on all three kinds: 3 figures, 3 toggles, and the POST round trip
blurs and un-blurs. 167 tests pass.

tests/test_booth.py

@@ -1447,3 +1447,43 @@ def test_blur_applies_to_inline_docs_not_just_images(client):
     assert body.count('<figure class="item item-doc blurred"') == 1, (
         "exactly the blurred doc, not every doc"
     )
+
+
+def test_every_item_kind_gets_exactly_one_blur_toggle(client):
+    """The regression guard for this whole feature's recurring bug.
+
+    booth.html has THREE item branches — doc, media, other — and each pass at
+    this feature patched some and missed others: first the blurred class landed
+    on media only, then the toggle landed on media only and the operator asked
+    "no UI option to blur/unblur?". A count assertion across mixed kinds is the
+    check that actually catches it; a spot check on one item never will.
+    """
+    c, root = client
+    d = root / "bo"
+    d.mkdir()
+    (d / "note.txt").write_text("doc branch")       # -> item-doc
+    _png(d / "pic.png")                              # -> item-image
+    (d / "blob.bin").write_bytes(b"\x00\x01binary")  # -> item-other
+
+    body = c.get("/b/bo/").text
+    figures = body.count('<figure class="item item-')
+    toggles = body.count('class="blurtoggle')
+    assert figures == 3, f"expected all three kinds to render, got {figures}"
+    assert toggles == figures, (
+        f"{figures} items but {toggles} blur toggles — a branch was missed again"
+    )
+
+
+def test_blur_toggle_posts_the_opposite_state(client):
+    """The button must flip, not just set. A toggle hard-coded to on=1 looks
+    identical in the markup and silently cannot un-blur."""
+    c, root = client
+    d = root / "bo"
+    d.mkdir()
+    _png(d / "pic.png")
+
+    assert 'name="on" value="1"' in c.get("/b/bo/").text
+    set_blurred(d, "pic.png", True)
+    body = c.get("/b/bo/").text
+    assert 'name="on" value="0"' in body, "a blurred item must offer un-blur"
+    assert "◉ blurred" in body